“Composite” Atlantic hurricane seasons

This is the general tropical discussion area. Anyone can take their shot at predicting a storms path.

Moderator: S2k Moderators

Forum rules

The posts in this forum are NOT official forecasts and should not be used as such. They are just the opinion of the poster and may or may not be backed by sound meteorological data. They are NOT endorsed by any professional institution or STORM2K. For official information, please refer to products from the National Hurricane Center and National Weather Service.

Help Support Storm2K
Message
Author
Shell Mound
Category 5
Category 5
Posts: 2434
Age: 31
Joined: Thu Sep 07, 2017 3:39 pm
Location: St. Petersburg, FL → Scandinavia

“Composite” Atlantic hurricane seasons

#1 Postby Shell Mound » Sun Sep 19, 2021 11:58 am

As has been noted, many discussions on this forum have sought to quantify how “extreme” an Atlantic hurricane season may realistically become, using sundry metrics such as seasonal ACE (long-tracking [major] hurricanes), longevity of activity (both early and late, pre and postseason), the number of landfalls and regions impacted, and so on. Looking back at climatology and history, and taking into account likely undercounts/-estimates prior to satellite and reconnaissance, I have pondered a realistic “ceiling” for Atlantic hurricane seasons. In doing so, I have resorted to NOAA’s Historical Hurricane Tracks (HHT) and developed “composite” seasons, selecting individual storms or “clusters” from a set of seasons. Based on my observations over the years, I have noticed that storms on a seasonal basis tend to concentrate in specific areas, follow certain trajectories, and adhere to certain patterns, in turn reflecting oceanic/atmospheric conditions that correspond to unique sets of parameters. I have done so while attempting to account for significant gaps in the official records prior to satellite and reconnaissance, including the fact that studies such as Landsea et al. (2004) and Vecchi et al. (2008) indicated ≥ 2.2 to 3.4 missing NS per year (possibly as many as six to eight per year) in or prior to 1965.

For the purpose at hand I have begun by looking at storms from the following six seasons: 1899, 1926, 1932, 1933, 1935, and 1941. Using NOAA’s HHT, I selected given storms from each of these seasons and thus formed a “composite” season. I was attempting to formulate a plausible “ceiling” for an Atlantic hurricane season that would be extreme but not implausible, especially over a long period of observation. For example, I looked at the 1933 Atlantic hurricane season and noted some key data. Officially, 1933 featured a seasonal ACE of 259—the highest on record since 1851—along with 20 NS, 11 H, and six MH. As noted in the previous link, data suggest that there may have been at least eight MH. Of the total NS, at least three (Storms #9, #19, and #20) featured MSW of ≥ 50 kt and estimated or observed MSLP of ≤ 1000 mb in areas of sparse data coverage, and so likely reached H status sometime during their lifespans. In particular, #19 may have even been a MH. Adding three new H, including one MH, and upgrading one existing H (Storm #2) to a MH, while adjusting for “missing” NS, would yield ~22–23 NS (if not more), 14 H, and 8 MH. Even these estimates are conservative in all three categories, and total NS could have been as many as ~26–28.

Applying the data to my methodology, for my first “test-case” I came up with the following “composite” hypothetical:

Image

Scenario #1: a season that is a blend of 1899, 1926, 1932, 1933, 1935, and 1941
 26 NS, 14 H, 11 MH
Highlights: The first NS forms on 5 May and the last NS dissipates on 14 November. Nine NS form in the MDR during the season, two of which become H there. Six H strike the CONUS, including three MH: two in FL and one in TX. Two MH strike Central America (inclusive of Yucatán), and two others affect Cuba. Of the 11 MH to form during the season, three peak at ≥ 140 kt (C5), two peak at ≥ 150 kt, and one peaks at ≥ 160 kt.
3 likes   
CVW / MiamiensisWx / Shell Mound
The posts in this forum are NOT official forecasts and should not be used as such. They are just the opinion of the poster and may or may not be backed by sound meteorological data. They are NOT endorsed by any professional institution or STORM2K. For official information, please refer to products from the NHC and NWS.

Shell Mound
Category 5
Category 5
Posts: 2434
Age: 31
Joined: Thu Sep 07, 2017 3:39 pm
Location: St. Petersburg, FL → Scandinavia

Re: “Composite” Atlantic hurricane seasons

#2 Postby Shell Mound » Sun Sep 19, 2021 2:11 pm

As an addendum to this, I would like to add some illuminating statistics and graphics from Klotzbach et al. (2021) to illustrate just how extreme the 1933 Atlantic hurricane season truly was.

Firstly, according to the study, in 1933 no fewer than eight storms tracked through the Caribbean while at hurricane intensity. This is a record in HURDAT, going back to 1851. Additionally, 1933 featured four H landfalls in Mexico (Storms #2, #5, #13, and #14), tying 2020’s record high (Nana, Delta, Gamma, and Zeta). 1933 caused severe flooding and wind damage from successive hurricane impacts on the Yucatán and Tampico, Tamaulipas. 1955 featured a very similar outcome for the same regions, with Hurricanes Gladys, Hilda, and Janet devastating these areas. Both 1933 and 1955 were very deadly years for Mexico, killing thousands of people due to hurricane-caused flash flooding via heavy rainfall, along with storm surge.

Like 2005, 1933 generated more ACE in the month of July than in the entire month of October, joining a very exclusive club. (1933, like 2005, also featured multiple, long-tracking hurricanes in the Caribbean during the month of July.) However, unlike 2005, 1933 produced the second-highest monthly ACE ever registered to date in the month of September (111 × 104 kt2), behind only 1926, which generated 132 × 104 kt2 in September. Since 1933, only 1961 (140 × 104 kt2), 2004 (155 × 104 kt2), and 2017 (174 × 104 kt2) have generated more ACE in September.

Image

According to Table 1 (above) from the study, 1933 featured 125.25 NS days (ranking: second) vs. 2005’s record-breaking 126.25 NS (ranking: first) days. Considering likely undercounts in the pre-satellite, pre-reconnaissance era, this is extremely impressive, and the difference between the two years is almost nonexistent despite sparser data coverage in 1933 vs. 2005. Even more astoundingly, 1933 solidly surpasses 2005 in terms of hurricane days, with 1933 featuring 57 (ranking: fourth) vs. 2005’s 49.75 (ranking: eighth). In terms of MH days, once again 1933 leads the pack, with 21.75 (ranking: fifth) vs. 2005’s 17.5 (ranking: seventh). And as I mentioned in my initial post in this thread, 1933’s actual NS, H, and MH days are likely substantially underestimated in each category!

Furthermore, the reanalysis project itself found that 1933 had 22% more seasonal ACE than originally indicated, thus increasing from 213 to 259. Even so, literature cited in the above study indicate that MSW for individual cyclones over the open sea between 1886 and 1943 were likely too low on the order of ≥ 10 to 20 kt, leading to substantial underestimates in ACE as well as intensity, especially in busy or hyperactive seasons such as 1933 that are heavily weighted toward the deep tropics (MDR).

Image

Figure 6 (above), extracted from the study, compares conditions during the 1933 and 2005 seasons. It clearly shows that 1933 actually featured lower MSLP and weaker vertical wind shear in the deep tropics, including the Caribbean and MDR. Some of this may be related to the fact that 1933 featured a 2005-type +AMO, yet coincided with a much colder tropical Pacific, as a moderate Niña episode was solidly in place in 1933, unlike in 2005. 1933 also featured less mid-level ridging (subsidence) in the deep tropics than 2005, hence a more unstable environment, especially in the MDR. Also, in 1933 SST in the Atlantic basin, especially the deep tropics (Caribbean and MDR), were much warmer than the rest of the global tropics—the fourth highest “relative” SST on record, in fact—thereby focussing rising air and vorticity over the Atlantic basin. According to the study, moreover, conditions during 1932–3—both seasons of which featured two or more Category-5s each—were similar to those during the years 1878, 1886–7, and 1893, all of which have been preliminarily reanalysed as extremely active season, with 1887 in particular featuring 24 NS.

The study also highlights that even reanalysed databases do not fully capture just how conducive conditions were in 1933 (and many other historic seasons prior to satellite and reconnaissance), since June “hind-casts” using these data only ended up yielding seasonal forecasts that called for an ACE of ~165, rather than ≥ 259. (August “hind-casts,” on the other hand, did better, calling for seasonal ACE of up to 228. Additionally, modelling using available databases only explain ≤ 40% of the variability in seasonal ACE between 1878 and 1932.

Early on 3 September Storms #8 and #11 were both 120-kt Category-4s nearing landfalls in South Texas and South Florida, respectively, while only ~700 n mi apart from each other. Both storms would go on to strike TX and FL as 110-kt major hurricanes—borderline Category-3/-4 systems—within 24h of each other. Never before or since have two MH struck the CONUS in such rapid succession.

The reanalysis of Storm #8 even acknowledges some uncertainty in regard to its intensity at landfall in TX, given that Brownsville recorded a minimum MSLP of 949 mb simultaneously with 70-kt NW winds, which, given an estimated RMW of 20 n mi (average: 16 n mi) and a slower-than-average forward speed, would suggest a MSLP as low as 935 mb over the coastline. So the MSW at LF could have been either 110 or 115 kt, but the TC may have either undergone or was undergoing an EWRC, so the reanalysis conservatively went with 110 kt. Even so, Brownsville was inland, on the southwestern side of the RMW, and Storm #8 was moving westward at the time, suggestive of deep-layer ridging and possibly low vertical wind shear, so I think one could easily argue for an intensity of 115 kt / 935 mb at LF, if not a bit stronger.

As far as Storm #11 is concerned, its smaller-than-average eye passed over or near Jupiter Inlet early on 4 September, while moving briskly past the local weather station, which registered a forty-minute-long “lull” (but not a dead calm) and a minimum MSLP of 948 mb. Almost exactly 18h earlier, the storm had made landfall on Harbour Island, Eleuthera, in the Bahamas, with observers there reporting a thirty-minute-long “calm” and a minimum MSLP of 945 mb. At the time the storm was also compact and moving rapidly, so the reanalysis assigned a MSW of 120 kt at LF on Eleuthera. Given that Jupiter, while inside the RMW, may have missed the exact wind-centre, the MSLP could have been a few mb lower than 948 mb at LF in FL, meaning that the storm’s intensity was essentially unchanged in the 18h interval between Eleuthera and Jupiter Inlet.

Another intriguing clue is that official descriptions of the damage in Martin County, just north of the eye, indicate that some relatively well-built homes were blown apart, while many CBS-style homes shifted off their masonry foundations. Fully three-fourths of the roofs in Stuart were severely damaged. However, the most severe damage occurred in the community of Olympia “Beach,” now a ghost town, located roughly along SE Olympus Street in present-day Hobe Sound. In this area numerous homes were destroyed and vegetation severely damaged, though oceanfront, wintertime estates, across the Indian River on the barrier island opposite Olympia and Hobe Sound, were better built less severely damaged. Overall, the damage better fits the SSHWS descriptor of Category-4 vs. Category-3 damage.

Given these reports and the official landfall location of 26.9°N, 80.1°W, along with a WNW movement placing the strongest winds NNE of the eye, the RMW at LF was likely no larger than 10 n mi (the reanalysis lists 15 n mi, which to me seems a bit large, though even this is smaller than the climatological average of 17 n mi). The MSLP at LF was likely close to the 945 mb that was measured 18h earlier in the Bahamas. A much-smaller-than-average RMW of ~10 n mi, coupled with a somewhat-smaller-than-average circulation and a faster-than-normal forward speed, would, when combined with an estimated MSLP of 945 mb at LF, lend a bit more support to 115 kt vs. 110 kt at LF in Florida, considering the essentially steady-state size/intensity between the Bahamas and Jupiter Inlet.

So, in all likelihood, 1933 featured two Category-4 landfalls (≥ 115 kt / ≤ 945 mb) on the CONUS within 24h of each other. Even measured against the statistics of years such as 2004 and 2017, that kind of incident is absolutely earthshaking.



Link: https://youtu.be/MqqmkqFfyms
6 likes   
CVW / MiamiensisWx / Shell Mound
The posts in this forum are NOT official forecasts and should not be used as such. They are just the opinion of the poster and may or may not be backed by sound meteorological data. They are NOT endorsed by any professional institution or STORM2K. For official information, please refer to products from the NHC and NWS.

tolakram
Admin
Admin
Posts: 19138
Age: 60
Joined: Sun Aug 27, 2006 8:23 pm
Location: Florence, KY (name is Mark)

Re: “Composite” Atlantic hurricane seasons

#3 Postby tolakram » Sun Sep 19, 2021 3:47 pm

Another intriguing clue is that official descriptions of the damage in Martin County, just north of the eye, indicate that some relatively well-built homes were blown apart, while many CBS-style homes shifted off their masonry foundations.


1933 homes built when? What was well built in 1933? Be careful with strength estimates in these older seasons. 1933 was a beast of a season, no doubt.
2 likes   
M a r k
- - - - -
Join us in chat: Storm2K Chatroom Invite. Android and IOS apps also available.

The posts in this forum are NOT official forecasts and should not be used as such. Posts are NOT endorsed by any professional institution or STORM2K.org. For official information and forecasts, please refer to NHC and NWS products.

Shell Mound
Category 5
Category 5
Posts: 2434
Age: 31
Joined: Thu Sep 07, 2017 3:39 pm
Location: St. Petersburg, FL → Scandinavia

Re: “Composite” Atlantic hurricane seasons

#4 Postby Shell Mound » Mon Sep 20, 2021 9:57 am

tolakram wrote:
Another intriguing clue is that official descriptions of the damage in Martin County, just north of the eye, indicate that some relatively well-built homes were blown apart, while many CBS-style homes shifted off their masonry foundations.

1933 homes built when? What was well built in 1933? Be careful with strength estimates in these older seasons. 1933 was a beast of a season, no doubt.

Prior to the postwar housing boom, and especially prior to the 1960s, construction standards in the coastal and interior South were either a) generally superior or b) relied on higher-quality materials than do prefabricated and/or mass-produced residences. For example, homebuilders sought the highest-quality timber for interior use. In South Florida, for instance, early residents extensively utilised mature “Dade County pine” with which to erect buildings. Even mansions were often built better, using such materials as oolite (“coral rock” or limestone) or stucco, along with the best available wood. That is why so many structures erected in the 1920s or earlier survived the inner cores of both the 1926 Miami hurricane and Andrew (1992) with only superficial damage, most of which, if any, was related to storm surge and rainfall seeping through shattered, unprotected windows. For the same reason antebellum—that is, pre-Civil War—as well as pre-World War I/II structures survived the inner core of Camille (1969) relatively well, but did not survive Katrina’s (2005) much higher storm surge years later. All this is well attested; for example, see the discussion by iCyclone, beginning at 34:50 in the following clip:



Link: https://youtu.be/jARp26WeLuU

Starting at 35:58 he even compares structures that survived Camille to “tanks” in terms of their construction, calling them “rock-solid,” including his very own “Hurricane House” in Pass Christian, which is a century-old cottage. Beginning at 36:38 he also notes that, while Andrew and Charley (2004) were indeed powerful hurricanes at landfall, the vast majority of the structures that they destroyed were either frail or of very dubious construction at best, despite being newly erected “modern houses” (his own descriptor at 36:35). Similar phenomena have been observed following intense tornadoes such as those at Gallatin, TN (2006), and Washington, IL (2013). Both these communities featured a large number of new, platted, “cookie-cutter” developments containing multi-story residential housing that was both a) very poorly secured and b) of low-quality material. In many cases even having a qualified structural engineer “on site” does not always give the full context that longtime residents and homebuilders themselves do (in many cases the residents and homebuilders are one and the same, primarily in older, more “rooted” communities, and hand down their occupations/stations from father to son, mother to daughter, generation after generation).

So, yes, 1933 was certainly a “beast” of a season, arguably more so than even 2005 or 2020, at least in certain respects. In addition to MH, one should not omit Storm #6, the notorious Chesapeake–Potomac hurricane. Despite being “only” a high-end Category-1 (80 kt / 963 mb) at landfall on the Outer Banks, its impacts to the Mid-Atlantic were arguably on a par with or, more frequently, exceeded those that were observed during Isabel (2003), in terms of maximum sustained wind, storm surge, rainfall, and widespread, severe freshwater flooding. Sustained winds of 71 kt at 10 m occurred in Virginia during the passage of the 1933 hurricane, whereas no station in the state measured sustained hurricane-force winds during Isabel, though both storms crossed portions of the state as hurricanes. As far as precipitation is concerned, Schoner et al. (1956, p. 184) indicate that 1933 #6 generated a maximum storm-total rainfall of 13.2”, with totals of ≥ 5” to 10” over a wide area, whereas 2003’s Isabel produced isolated pockets of ≥ 10”, including a small, isolated maximum of 20.2”. In fact, 1933 #6 generated more severe flooding over a wider area than any other storm regionally except 1972’s Agnes.

Per Klotzbach et al. (2021), 1933 #6 would likely cause ≥ $30–38 billion in losses today, adjusted for inflation, and inflation-adjusted estimates for most historical storms tend to be conservative.

Other highlights of the 1933 season:

  • 1933 #12, the Outer Banks hurricane, produced sustained winds of 79 kt NW at Cape Hatteras, which experienced the western portion of the eye, and those winds were extrapolated, given that the anemometer was partly disabled prior to the worst conditions, after recording 66-kt winds. The storm featured a somewhat large RMW of ~35 n mi or possibly twin wind maxima as it passed just offshore, hence the very high storm tides of ~1.2 m (~4 ft) at New Bern, NC, far from the eye, where 100+ trees of ≥ 10” (roughly a foot in diameter!) in diameter were downed by strong winds, although this area experienced little rainfall from Storm #6. So this storm was likely stronger than Irene along the Outer Banks and in other portions of eastern North Carolina, despite being similar in size.
2 likes   
CVW / MiamiensisWx / Shell Mound
The posts in this forum are NOT official forecasts and should not be used as such. They are just the opinion of the poster and may or may not be backed by sound meteorological data. They are NOT endorsed by any professional institution or STORM2K. For official information, please refer to products from the NHC and NWS.

tolakram
Admin
Admin
Posts: 19138
Age: 60
Joined: Sun Aug 27, 2006 8:23 pm
Location: Florence, KY (name is Mark)

Re: “Composite” Atlantic hurricane seasons

#5 Postby tolakram » Mon Sep 20, 2021 10:22 am

I don't have a strong argument against any of that but there are biases (let's call them romantic notions of the past) that need to be accounted for. Infrastructure built today is a lot different from that in the 30's. Old pines are fine, I have no doubt many of these homes were built with stronger materials, but I doubt they had roof or foundation strapping, or some of the other small improvements that can hold even a less well built house together. While a house might have been structurally superior to todays homes it only takes a roof ripping apart to bring down an entire structure. Any analysis of this? My google skills failed me looking up when roofing / foundation straps or bracing was first used or commonly used.
1 likes   
M a r k
- - - - -
Join us in chat: Storm2K Chatroom Invite. Android and IOS apps also available.

The posts in this forum are NOT official forecasts and should not be used as such. Posts are NOT endorsed by any professional institution or STORM2K.org. For official information and forecasts, please refer to NHC and NWS products.

Shell Mound
Category 5
Category 5
Posts: 2434
Age: 31
Joined: Thu Sep 07, 2017 3:39 pm
Location: St. Petersburg, FL → Scandinavia

Re: “Composite” Atlantic hurricane seasons

#6 Postby Shell Mound » Wed Sep 22, 2021 2:02 pm

tolakram wrote:I don't have a strong argument against any of that but there are biases (let's call them romantic notions of the past) that need to be accounted for. Infrastructure built today is a lot different from that in the 30's. Old pines are fine, I have no doubt many of these homes were built with stronger materials, but I doubt they had roof or foundation strapping, or some of the other small improvements that can hold even a less well built house together. While a house might have been structurally superior to todays homes it only takes a roof ripping apart to bring down an entire structure. Any analysis of this? My google skills failed me looking up when roofing / foundation straps or bracing was first used or commonly used.

I am not a structural engineer or an architect, but I think that the distribution of weight was more even in older, well-built structures. Plus, many of these structures were designed to account for aerodynamics, incorporating sloping, “hipped” roofs, relatively “open” plans, and numerous interior walls that, combined with high-quality materials, added to overall wind-resistance and negated the need for “strapping.” Structures were simply heavier on average, yet less “top-heavy” than today’s, given greater volume. In other words, a violent tornado would more easily demolish a “modern” bolted-and-strapped and/or prefabricated residence than, say, an antebellum, plantation-style mansion. Nevertheless, as acknowledged previously, I am certainly not a qualified expert, so others are certainly free, without qualification, to “weigh in” on this matter.
0 likes   
CVW / MiamiensisWx / Shell Mound
The posts in this forum are NOT official forecasts and should not be used as such. They are just the opinion of the poster and may or may not be backed by sound meteorological data. They are NOT endorsed by any professional institution or STORM2K. For official information, please refer to products from the NHC and NWS.

User avatar
Ubuntwo
Category 5
Category 5
Posts: 1089
Age: 30
Joined: Thu Oct 05, 2017 10:41 pm

Re: “Composite” Atlantic hurricane seasons

#7 Postby Ubuntwo » Wed Sep 22, 2021 3:52 pm

Shell Mound wrote:
As far as Storm #11 is concerned, its smaller-than-average eye passed over or near Jupiter Inlet early on 4 September, while moving briskly past the local weather station, which registered a forty-minute-long “lull” (but not a dead calm) and a minimum MSLP of 948 mb. Almost exactly 18h earlier, the storm had made landfall on Harbour Island, Eleuthera, in the Bahamas, with observers there reporting a thirty-minute-long “calm” and a minimum MSLP of 945 mb. At the time the storm was also compact and moving rapidly, so the reanalysis assigned a MSW of 120 kt at LF on Eleuthera. Given that Jupiter, while inside the RMW, may have missed the exact wind-centre, the MSLP could have been a few mb lower than 948 mb at LF in FL, meaning that the storm’s intensity was essentially unchanged in the 18h interval between Eleuthera and Jupiter Inlet.

Another intriguing clue is that official descriptions of the damage in Martin County, just north of the eye, indicate that some relatively well-built homes were blown apart, while many CBS-style homes shifted off their masonry foundations. Fully three-fourths of the roofs in Stuart were severely damaged. However, the most severe damage occurred in the community of Olympia “Beach,” now a ghost town, located roughly along SE Olympus Street in present-day Hobe Sound. In this area numerous homes were destroyed and vegetation severely damaged, though oceanfront, wintertime estates, across the Indian River on the barrier island opposite Olympia and Hobe Sound, were better built less severely damaged. Overall, the damage better fits the SSHWS descriptor of Category-4 vs. Category-3 damage.

Given these reports and the official landfall location of 26.9°N, 80.1°W, along with a WNW movement placing the strongest winds NNE of the eye, the RMW at LF was likely no larger than 10 n mi (the reanalysis lists 15 n mi, which to me seems a bit large, though even this is smaller than the climatological average of 17 n mi). The MSLP at LF was likely close to the 945 mb that was measured 18h earlier in the Bahamas. A much-smaller-than-average RMW of ~10 n mi, coupled with a somewhat-smaller-than-average circulation and a faster-than-normal forward speed, would, when combined with an estimated MSLP of 945 mb at LF, lend a bit more support to 115 kt vs. 110 kt at LF in Florida, considering the essentially steady-state size/intensity between the Bahamas and Jupiter Inlet.


Eye pressure gradients are pretty shallow, 1mb/10kt. Jupiter recorded a 40 minute calm (longer than in the Bahamas), so I'm not sure how landfall could be much deeper than 948. KZC analyses for that pressure/size/latitude come out to 110kt.

Stuart/Hobe Sound was not mostly "antebellum plantation-style housing". Going off old photos, solid construction was limited to downtown.
Here is Hobe Sound's train station in the 60s:
Image

Stuart neighborhood in 1918. All wood houses. We have seen this type of structure destroyed in category 1 wind.
Image

Wooden structures in downtown Stuart, 1926.
Image

About as good as it gets, also 1926. Hurricane Hugo produced 87 mph winds in Charleston and destroyed many similar structures. See also Michael in Marianna FL.
Image

'Filling in the blanks' with storms is hard. To play devil's advocate, many recent storms would probably be rated higher if they occurred earlier. Katrina made landfall at 920 mb with a 30 foot surge. Without recon, how is that anything other than a 5? And a pretty strong one at that. Ike and Florence had landfall MSLP typical of a 3. Rita at 937 mb, surely that must be a cat 4? Ivan at 943 mb, Dennis at 946 mb, surely all 4s as well?
Not to mention storms with inefficient mixing, where SFMR kept the NHC from a higher intensity. Humberto, Isaias, and Zeta are good recent examples.

That isn't to diminish your point. 1933 was absolutely the most active hurricane season on record, and I imagine plenty of storms were underestimated. Just probably not the ones landfalling in observation-rich slices of the US coast.
3 likes   
Kendall -> SLO -> PBC

Memorable Storms: Katrina (for its Florida landfall...) Wilma Matthew Irma

Shell Mound
Category 5
Category 5
Posts: 2434
Age: 31
Joined: Thu Sep 07, 2017 3:39 pm
Location: St. Petersburg, FL → Scandinavia

Re: “Composite” Atlantic hurricane seasons

#8 Postby Shell Mound » Thu Sep 23, 2021 2:31 pm

Ubuntwo wrote:
Shell Mound wrote:As far as Storm #11 is concerned, its smaller-than-average eye passed over or near Jupiter Inlet early on 4 September, while moving briskly past the local weather station, which registered a forty-minute-long “lull” (but not a dead calm) and a minimum MSLP of 948 mb. Almost exactly 18h earlier, the storm had made landfall on Harbour Island, Eleuthera, in the Bahamas, with observers there reporting a thirty-minute-long “calm” and a minimum MSLP of 945 mb. At the time the storm was also compact and moving rapidly, so the reanalysis assigned a MSW of 120 kt at LF on Eleuthera. Given that Jupiter, while inside the RMW, may have missed the exact wind-centre, the MSLP could have been a few mb lower than 948 mb at LF in FL, meaning that the storm’s intensity was essentially unchanged in the 18h interval between Eleuthera and Jupiter Inlet.

Another intriguing clue is that official descriptions of the damage in Martin County, just north of the eye, indicate that some relatively well-built homes were blown apart, while many CBS-style homes shifted off their masonry foundations. Fully three-fourths of the roofs in Stuart were severely damaged. However, the most severe damage occurred in the community of Olympia “Beach,” now a ghost town, located roughly along SE Olympus Street in present-day Hobe Sound. In this area numerous homes were destroyed and vegetation severely damaged, though oceanfront, wintertime estates, across the Indian River on the barrier island opposite Olympia and Hobe Sound, were better built less severely damaged. Overall, the damage better fits the SSHWS descriptor of Category-4 vs. Category-3 damage.

Given these reports and the official landfall location of 26.9°N, 80.1°W, along with a WNW movement placing the strongest winds NNE of the eye, the RMW at LF was likely no larger than 10 n mi (the reanalysis lists 15 n mi, which to me seems a bit large, though even this is smaller than the climatological average of 17 n mi). The MSLP at LF was likely close to the 945 mb that was measured 18h earlier in the Bahamas. A much-smaller-than-average RMW of ~10 n mi, coupled with a somewhat-smaller-than-average circulation and a faster-than-normal forward speed, would, when combined with an estimated MSLP of 945 mb at LF, lend a bit more support to 115 kt vs. 110 kt at LF in Florida, considering the essentially steady-state size/intensity between the Bahamas and Jupiter Inlet.


Eye pressure gradients are pretty shallow, 1mb/10kt. Jupiter recorded a 40 minute calm (longer than in the Bahamas), so I'm not sure how landfall could be much deeper than 948. KZC analyses for that pressure/size/latitude come out to 110kt.

As I explained above, the semantics indicate that “calm” conditions were reported in the Bahamas, while a “decided lull” (Florida Climatological Data, September 1933, special section) occurred in South Florida. “Calm” implies an absence or near-absence of discernible wind, whereas “lull” denotes a relative rather than absolute decrease in velocities. The reports from the Bahamas did not specify whether the thirty-minute period of “calm” coincided with a longer “lull.” One may surmise that the very centre of the eye passed overhead, providing a period of absolute calm, but that this calm occurred within a longer lull provided by the passage of the RMW (eye) itself. By contrast, the wind only decreased for a forty-minute period at Jupiter, but did not reach a calm state, so arguably the eye’s/RMW’s passage at Jupiter was briefer than at Harbour Island, Eleuthera. Storm #11 was a compact system in its passage through the Bahamas and South Florida, so the gradient inside the eye/RMW would have been steeper than usual. Also, the Storm #4 of 1947, the Fort Lauderdale hurricane, featured a gradient of ~1 mb/n mi inside the RMW, based on numerous observations: for example, downtown Fort Lauderdale reported 956 mb and a “lull” of 1h 15min, while Hillsboro Inlet Light, about ~9 n mi distant, registered 947 mb and a “lull” of perhaps half an hour, near the northern edge of the eye.
Ubuntwo wrote:Stuart/Hobe Sound was not mostly "antebellum plantation-style housing".

I never said that it was.
Ubuntwo wrote:Going off old photos, solid construction was limited to downtown.
Here is Hobe Sound's train station in the 60s:
https://i.imgur.com/taccZZK.png

Stuart neighborhood in 1918. All wood houses. We have seen this type of structure destroyed in category 1 wind.
https://i.imgur.com/BvU2LWe.png

Façades and porches, certainly. The structures themselves? No.

Ubuntwo wrote:About as good as it gets, also 1926. Hurricane Hugo produced 87 mph winds in Charleston and destroyed many similar structures. See also Michael in Marianna FL.
https://i.imgur.com/s8SsAWP.png

“Many similar structures”? Hugo did destroy the upper stories of unreinforced brick structures in Charleston. However, most other structures, especially historic, survived with superficial damage, except those that contained tin or tile roofing. Also, as mentioned, Hugo generated sustained winds of 76 kt in Charleston, on the weaker side of the eye. As I mentioned above, Stuart was actually ~10 n mi north of the RMW, which affected the settlement of Olympia (Beach) in present-day Hobe Sound.

Ubuntwo wrote:'Filling in the blanks' with storms is hard. To play devil's advocate, many recent storms would probably be rated higher if they occurred earlier. Katrina made landfall at 920 mb with a 30 foot surge. Without recon, how is that anything other than a 5? And a pretty strong one at that. Ike and Florence had landfall MSLP typical of a 3. Rita at 937 mb, surely that must be a cat 4? Ivan at 943 mb, Dennis at 946 mb, surely all 4s as well?
Not to mention storms with inefficient mixing, where SFMR kept the NHC from a higher intensity. Humberto, Isaias, and Zeta are good recent examples.

Following Drs. Saffir and Simpson’s lead, here is an overview of the old SSHS categories, as distinguished by minimum central pressure:

Category 1: ≥980 mb
Category 2: 965-79 mb
Category 3: 945-64 mb
Category 4: 920-44 mb
Category 5: ≤919 mb

By this metric, Katrina’s 920 mb would classify the storm as a high-end “4” on the old, pressure-based scale, not a “pretty strong” Cat-5! :wink:

Regardless, this final paragraph is more incidental than central to the specific points that I have raised in relation to very specific storms and circumstances.
0 likes   
CVW / MiamiensisWx / Shell Mound
The posts in this forum are NOT official forecasts and should not be used as such. They are just the opinion of the poster and may or may not be backed by sound meteorological data. They are NOT endorsed by any professional institution or STORM2K. For official information, please refer to products from the NHC and NWS.

Shell Mound
Category 5
Category 5
Posts: 2434
Age: 31
Joined: Thu Sep 07, 2017 3:39 pm
Location: St. Petersburg, FL → Scandinavia

Re: “Composite” Atlantic hurricane seasons

#9 Postby Shell Mound » Thu Sep 23, 2021 3:01 pm

Regarding historic seasons, thanks to the esteemed efforts of TheAustinMan, I have decided to post the following charts highlighting the preliminary “best track” (see the Excel files under “Supplementary materials”) for a select number of seasons, based on the reanalysis linked previously. As can be seen, the preliminary reanalysis found that these and other seasons were far more active and intense than officially recognised, especially in the MDR and subtropics. Simply compare the following charts from TheAustinMan to the current HURDAT (note that the colour-scheme is the same as Wikipedia’s for respective intensities):

Image
Image
Image
Image
Image
Image
Image

Some highlights:
  • 1886 is like a much more severe version of 2008, heavily west-based and Gulf-centred but with several CV systems as well.
  • 1887 featured no fewer than 24 NS. Given observational limitations, both it and 1933 likely matched or exceeded 2005’s total NS.
  • 1893 featured no fewer than six hurricanes in the MDR, if not more. All those Cat-3s were likely Cat-4+ over the Atlantic.
  • Five of the seven seasons—all but 1870 and 1878—featured four or more major hurricanes, as conservatively estimated by the reanalysis.
  • Four-fifths, if not all, of the seasons featured multiple hurricanes in the MDR. The two exceptions likely did so as well.

Bottom line: science absolutely needs to delve more deeply into widely dispersed historical records, on a truly international scale, and conduct thorough reanalyses of all major basins—not only the Atlantic, but also the Pacific and Indian basins. Satellite imagery already indicates that storms from the latter two basins have been massively underestimated from the 1960s through the early 1990s. If a thorough, global reanalysis were conducted, much, if not all, of the estimated increase in MPI due to climate change might disappear. Although climate change is real, the atmosphere, like the planet Earth herself, is quite complex. Perhaps there is a built-in equilibrium that keeps MPI in a relatively steady state, regardless of changes in SST and OHC/TCHP due to global warming.
Last edited by Shell Mound on Fri Sep 24, 2021 10:25 am, edited 1 time in total.
3 likes   
CVW / MiamiensisWx / Shell Mound
The posts in this forum are NOT official forecasts and should not be used as such. They are just the opinion of the poster and may or may not be backed by sound meteorological data. They are NOT endorsed by any professional institution or STORM2K. For official information, please refer to products from the NHC and NWS.

tolakram
Admin
Admin
Posts: 19138
Age: 60
Joined: Sun Aug 27, 2006 8:23 pm
Location: Florence, KY (name is Mark)

Re: “Composite” Atlantic hurricane seasons

#10 Postby tolakram » Thu Sep 23, 2021 5:01 pm

Shell Mound wrote:
... snip ...

Bottom line: science absolutely needs to delve more deeply into widely dispersed historical records, on a truly international scale, and conduct thorough reanalyses of all major basins—not only the Atlantic, but also the Pacific and Indian basins. Satellite imagery already indicates that storms from the latter two basins have been massively underestimated from the 1960s through the early 1990s. If a thorough, global reanalysis were conducted, much, if not all, of the estimated increase in MPI due to climate change might disappear. Although climate change is real, the atmosphere, like the planet Earth herself, is quite complex. Perhaps there is a built-in equilibrium that keeps MPI in a relatively steady state, regardless of changes in SST and OHC/TCHP due to global warming.


Why? What benefit will this have?

You're parsing words from the 1800's. I just don't get what you're after here. You seem to believe that storms in the 1800's were stronger? Or just as strong?
2 likes   
M a r k
- - - - -
Join us in chat: Storm2K Chatroom Invite. Android and IOS apps also available.

The posts in this forum are NOT official forecasts and should not be used as such. Posts are NOT endorsed by any professional institution or STORM2K.org. For official information and forecasts, please refer to NHC and NWS products.

User avatar
Ubuntwo
Category 5
Category 5
Posts: 1089
Age: 30
Joined: Thu Oct 05, 2017 10:41 pm

Re: “Composite” Atlantic hurricane seasons

#11 Postby Ubuntwo » Thu Sep 23, 2021 5:47 pm

Shell Mound wrote:As I explained above, the semantics indicate that “calm” conditions were reported in the Bahamas, while a “decided lull” (Florida Climatological Data, September 1933, special section) occurred in South Florida. “Calm” implies an absence or near-absence of discernible wind, whereas “lull” denotes a relative rather than absolute decrease in velocities. The reports from the Bahamas did not specify whether the thirty-minute period of “calm” coincided with a longer “lull.” One may surmise that the very centre of the eye passed overhead, providing a period of absolute calm, but that this calm occurred within a longer lull provided by the passage of the RMW (eye) itself. By contrast, the wind only decreased for a forty-minute period at Jupiter, but did not reach a calm state, so arguably the eye’s/RMW’s passage at Jupiter was briefer than at Harbour Island, Eleuthera. Storm #11 was a compact system in its passage through the Bahamas and South Florida, so the gradient inside the eye/RMW would have been steeper than usual. Also, the Storm #4 of 1947, the Fort Lauderdale hurricane, featured a gradient of ~1 mb/n mi inside the RMW, based on numerous observations: for example, downtown Fort Lauderdale reported 956 mb and a “lull” of 1h 15min, while Hillsboro Inlet Light, about ~9 n mi distant, registered 947 mb and a “lull” of perhaps half an hour, near the northern edge of the eye.

I'm guessing the semantic difference is distinct observers describing the same phenomenon with different words. There's no way to say for sure either way.

Regarding the Fort Lauderdale hurricane. The Fort Lauderdale coast guard station recorded 951 mb 14 miles from Lighthouse Point (945 mb, both of these obs are in HURDAT metadata). That would be .4 mb/mile within the eye.

Ubuntwo wrote:Stuart/Hobe Sound was not mostly "antebellum plantation-style housing".

I never said that it was.

Wasn't claiming you were. You do say construction was superior to justify a higher intensity. But if the best construction was in beachside homes on a barrier island in the RMW, experiencing relatively minimal damage, that doesn't really support cat 4 alone.

Ubuntwo wrote:Stuart neighborhood in 1918. All wood houses. We have seen this type of structure destroyed in category 1 wind.

Façades and porches, certainly. The structures themselves? No.

Pay SELA or Lake Charles suburbs a visit. Countless homes without roofs in areas recieving cat 1 sustained wind w/ higher gusts. Much of the damage described in this Treasure Coast hurricane is to roofing.

“Many similar structures”? Hugo did destroy the upper stories of unreinforced brick structures in Charleston. However, most other structures, especially historic, survived with superficial damage, except those that contained tin or tile roofing. Also, as mentioned, Hugo generated sustained winds of 76 kt in Charleston, on the weaker side of the eye. As I mentioned above, Stuart was actually ~10 n mi north of the RMW, which affected the settlement of Olympia (Beach) in present-day Hobe Sound.

Point is, category 1 wind is capable of destroying brick structures. Who knows if the brick structures in Stuart were 'reinforced' or not.

Following Drs. Saffir and Simpson’s lead, here is an overview of the old SSHS categories, as distinguished by minimum central pressure:

Category 1: ≥980 mb
Category 2: 965-79 mb
Category 3: 945-64 mb
Category 4: 920-44 mb
Category 5: ≤919 mb

By this metric, Katrina’s 920 mb would classify the storm as a high-end “4” on the old, pressure-based scale, not a “pretty strong” Cat-5! :wink:

You got me! But Buras recorded 920 mb, and only had a lull, so the pressure was probably a couple mb lower :P

Regardless, this final paragraph is more incidental than central to the specific points that I have raised in relation to very specific storms and circumstances.

I guess this wasn't super clear. My point is making assumptions w/ limited data can justify a lot. We will never know how many tropical storms formed in 1933. Nobody knows if a recon plane would find >113kt SFMR (or far lower) in this hurricane at landfall, and again, we never will. BT has only been solid for around 30 years, and knowledge of tropical cyclone dynamics and observation are only getting better. Maybe we should be looking forward?
2 likes   
Kendall -> SLO -> PBC

Memorable Storms: Katrina (for its Florida landfall...) Wilma Matthew Irma

User avatar
Blinhart
Category 5
Category 5
Posts: 1981
Age: 47
Joined: Thu Sep 08, 2011 3:13 pm
Location: Crowley, La.

Re: “Composite” Atlantic hurricane seasons

#12 Postby Blinhart » Thu Sep 23, 2021 11:44 pm

I'm sorry there is no way you can scientifically compare storm seasons from prior to 1900, to storms between 1900-1960, to storms between 1960-1990, to the current past 30 years. The studying of storms in each of these different time frames have way different ways of naming, categorizing, analyzing any of the systems. You don't know how many storms were missed, misidentified, or misinterpreted. So I don't know why anyone would actually do papers actually comparing these years.
4 likes   
Personal Forecast Disclaimer:
The posts in this forum are NOT official forecast and should not be used as such. They are just the opinion of the poster and may or may not be backed by sound meteorological data. They are NOT endorsed by any professional institution or storm2k.org. For official information, please refer to the NHC and NWS products.

Shell Mound
Category 5
Category 5
Posts: 2434
Age: 31
Joined: Thu Sep 07, 2017 3:39 pm
Location: St. Petersburg, FL → Scandinavia

Re: “Composite” Atlantic hurricane seasons

#13 Postby Shell Mound » Sun Oct 24, 2021 12:28 pm

In terms of seasonal ACE, 1870–1, 1878, 1887, 1893, and 1926 could very well have rivalled or exceeded either 2005 or 1933 or both.
2 likes   
CVW / MiamiensisWx / Shell Mound
The posts in this forum are NOT official forecasts and should not be used as such. They are just the opinion of the poster and may or may not be backed by sound meteorological data. They are NOT endorsed by any professional institution or STORM2K. For official information, please refer to products from the NHC and NWS.

User avatar
Ptarmigan
Category 5
Category 5
Posts: 5272
Joined: Wed Aug 16, 2006 9:06 pm

Re: “Composite” Atlantic hurricane seasons

#14 Postby Ptarmigan » Sun Oct 24, 2021 12:41 pm

I would not be surprised if some seasons do exceed 2005 and even 1933. I have suspected 1878, 1886, 1887, 1893, 1899, and 1926 being more active.
3 likes   

Shell Mound
Category 5
Category 5
Posts: 2434
Age: 31
Joined: Thu Sep 07, 2017 3:39 pm
Location: St. Petersburg, FL → Scandinavia

Re: “Composite” Atlantic hurricane seasons

#15 Postby Shell Mound » Tue Oct 26, 2021 11:47 am

Ptarmigan wrote:I would not be surprised if some seasons do exceed 2005 and even 1933. I have suspected 1878, 1886, 1887, 1893, 1899, and 1926 being more active.

I also wouldn’t exclude 1870–1 from the list. Reanalysis located at least three hurricanes in the MDR in 1870 and at least two in 1871. In addition, in 1870 there were three other systems that developed in the MDR and were either a) long-lived tropical storms or b) systems that became hurricanes while just north of the MDR; in 1871 there were four such systems. Given known observational deficiencies prior to satellite and reconnaissance, both 1870 and 1871 probably featured at least six or seven hurricanes in the MDR, including at least a few or three majors. Most of the aforementioned systems, all of which were long-lived, curved OTS or briefly impacted the Canadian Maritimes or Bermuda, so their peak intensities, even following reanalysis, are likely considerably underestimated. 1878, 1887, and 1893 appear to have been roughly as active in the MDR as 1870–1, if not more so. Along with 1926, 1870–1, 1878, 1887, and 1893 likely rivalled 2005 and/or 1933 in terms of either ACE or total NS or both.
2 likes   
CVW / MiamiensisWx / Shell Mound
The posts in this forum are NOT official forecasts and should not be used as such. They are just the opinion of the poster and may or may not be backed by sound meteorological data. They are NOT endorsed by any professional institution or STORM2K. For official information, please refer to products from the NHC and NWS.

User avatar
Ptarmigan
Category 5
Category 5
Posts: 5272
Joined: Wed Aug 16, 2006 9:06 pm

Re: “Composite” Atlantic hurricane seasons

#16 Postby Ptarmigan » Mon Nov 01, 2021 7:05 pm

Shell Mound wrote:
Ptarmigan wrote:I would not be surprised if some seasons do exceed 2005 and even 1933. I have suspected 1878, 1886, 1887, 1893, 1899, and 1926 being more active.

I also wouldn’t exclude 1870–1 from the list. Reanalysis located at least three hurricanes in the MDR in 1870 and at least two in 1871. In addition, in 1870 there were three other systems that developed in the MDR and were either a) long-lived tropical storms or b) systems that became hurricanes while just north of the MDR; in 1871 there were four such systems. Given known observational deficiencies prior to satellite and reconnaissance, both 1870 and 1871 probably featured at least six or seven hurricanes in the MDR, including at least a few or three majors. Most of the aforementioned systems, all of which were long-lived, curved OTS or briefly impacted the Canadian Maritimes or Bermuda, so their peak intensities, even following reanalysis, are likely considerably underestimated. 1878, 1887, and 1893 appear to have been roughly as active in the MDR as 1870–1, if not more so. Along with 1926, 1870–1, 1878, 1887, and 1893 likely rivalled 2005 and/or 1933 in terms of either ACE or total NS or both.


I would not be surprised if 1870 and 1871 were active. It looks like 1870s is an active decade.

I even think were even more active seasons prior before 19th century.

Many storms go undetected and some records are lost somewhere.
0 likes   


Return to “Talkin' Tropics”

Who is online

Users browsing this forum: WAcyclone, WoodberryWx and 68 guests